EP0486572B1

EP0486572B1 - Improvements in granulocyte-macrophage colony-stimulating factor receptor and derivatives thereof

Info

Publication number: EP0486572B1
Application number: EP90912329A
Authority: EP
Inventors: Nicos Anthony Nicola; Nicholas Martin Gough; David Paul Gearing; Donald Metcalf; Julie Ann King
Original assignee: Amrad Corp Ltd
Current assignee: CSL Innovation Pty Ltd
Priority date: 1989-08-11
Filing date: 1990-08-10
Publication date: 1998-01-07
Anticipated expiration: 2010-08-10
Also published as: US5726036A; JPH04507344A; JP3267966B2; AU637133B2; SG59954A1; CA2064814C; DK0486572T3; ATE161883T1; EP0486572A4; DE69031914T2; EP0486572A1; US6136957A; AU6189690A; DE69031914D1; CA2064814A1; WO1991002063A1

Abstract

The invention provides a recombinant or synthetic receptor for granulocyte-macrophage colony-stimulating factor (GM-CSF) and biochemical and/or biological equivalents, homologues, or derivatives thereof. The GM-CSF receptor is useful inter alia in treatment and diagnosis of GM-CSF related diseases, and in screening of agonists and antagonists of binding of GM-CSF to its cell-bound receptors. The invention further provides nucleic acid sequences encoding GM-CSF receptor, recombinant DNA molecules comprising these sequences, and transformed hosts carrying the recombinant DNA molecules, as well as antibodies directed against GM-CSF receptor, and pharmaceutical compositions comprising either the GM-CSF receptor or antibody thereto.

Description

The present invention relates generally to human recombinant and synthetic granulocyte-macrophage colony stimulating factor (GM-CSF) receptor and to biochemical and/or biological equivalents, homologues or derivatives thereof. These molecules are useful inter alia in the preparation of therapeutics and diagnostics and in the generation of agonists and antagonists with respect to the binding of GM-CSF to its receptor.

Granulocyte-macrophage colony-stimulating factor (GM-CSF) is a glycoprotein growth and differentiation factor which regulates the proliferation, differentiation and functional activity of cells of the neutrophil, eosinophil and monocyte/macrophage series (reviewed in Metcalf, 1984, and Gough and Nicola, 1989). Molecular clones encoding murine (Gough et el, 1984) and human (Wong et al, 1985) GM-CSF have been isolated, and recombinant protein tested in animal model systems (Metcalf et al; 1987, Donahue et al, 1986) and in phase I/II clinical trials in patients with a variety of haemopoietic disorders (reviewed in Morstyn et al, 1989). In both animal experiments and clinical trials, GM-CSF has been found to elevate circulating levels of monocytes, neutrophils and eosinophils; to enhance the functional capacities of the circulating cells; and to enhance the rate of haemopoietic recovery following chemotherapy and/or bone marrow transplantation (Gough and Nicola, 1989; Morstyn et al, 1989).

In both murine and human systems, autoradiographic analyses have indicated that GM-CSF receptors are present in low numbers (a few hundred per cell) on cells within the monocyte, neutrophil and eosinophil lineages (Nicola, 1987; DiPersio et al, 1988). However, functional GM-CSF receptors have also been detected on non-haemopoietic cells, including endothelial cells (Bussolino et al, 1989), small-cell lung carcinoma cell lines and SV40-transformed simian COS cells (Cocita Baldwin et al, 1989).

In the murine system, Walker and Burgess (1985) detected both high-affinity (K_D about 30pM) and low-affinity (K_D about lnM) receptors, while Park et al (1986a) detected a single receptor class of K_D 1-3nM. In the human system, only high-affinity receptors have been described on haemopoietic and endothelial cells (K_D about 30pM) (Gasson et al, 1986; Bussolino et al, 1989; Park et al, 1986b), but a receptor of lower affinity has been described on COS cells (Cocita Baldwin et al, 1989). In the murine system, the GM-CSF receptor(s) recognize only GM-CSF, although they can be indirectly down-modulated by interleukin-3 and other agents (Walker et al, 1985; Nicola, 1987).

Because most of the biological effects of GM-CSF are observed at picomolar concentrations (Metcalf, 1984), and because we have found that high-affinity receptors are preferentially internalized, Gough and Nicola, 1989), it is not clear whether low-affinity receptors are biologically functional. In the murine system at 37°C (Walker et al, 1985) and in the human system at 4°C or 37°C (Elliott et al, 1989; Lopez et al, 1989; Park et al, 1989), Multi-CSF (interkeukin-3) can down-modulate GM-CSF receptors on some types of hemopoietic cells, but it is not clear whether this is mediated by different receptor subclasses or by receptor-receptor interactions (Gearing et al, 1989; Elliot et al, 1989; Lopez et al, 1989).

GM-CSF was originally defined by its ability to stimulate the proliferation of granulocyte/macrophage progenitor cells, but more recently it has become apparent that it can also stimulate the proliferation of progenitor cells of other haemopoietic lineages (Metcalf et al, 1980) and cells of non-haemopoietic origin. The latter include human bone marrow fibroblasts, osteogenic sarcoma cell lines and a breast carcinoma cell line (Dedhar et al, 1988), human small cell carcinoma cell lines (Cocita Baldwin et al, 1989), human endothelial cells (Bussolino et al, 1989) and human placental cells (Wegmann et al, 1989). In addition, GM-CSF stimulates human endothelial cell migration (Bussolino et al, 1989), the proliferation and function of human osteoblast-like cells in vitro (Evans et al, 1989) and improves the growth of murine placental cells in vivo (Wegmann et al, 1989).

In view of these biological results, it is apparent that the GM-CSF receptors detected on these cells are functional, despite the fact that only high affinity receptors were detected on the endothelial cells, whereas only low affinity receptors were detected on the fibroblasts and placental membranes.

On haemopoietic cells, low affinity hGM-CSF receptors were distinguished by a rapid rate of ligand dissociation and poor internalization at 37°C, whereas high affinity receptors displayed a much slower rate of ligand dissociation, and were efficiently internalized. In addition to this complexity, two types of high affinity hGM-CSF receptor have been described on some but not all normal haemopoietic cells and cell lines. One type recognizes only hGM-CSF and is the only type of GM-CSF receptor on human neutrophils, while the other type apparently recognizes hGM-CSF and h-IL-3 with nearly equal affinity, and represents 80% of GM-CSF receptors on eosinophils (Lopez et al., 1989). Reciprocally, IL-3-specific or cross-reactive receptors have also been described (Park et al., 1989).

Finally, cross-linking experiments have suggested a molecular weight for the murine GM-CSF receptor of 51,000 (Walker and Burgess, 1985) or 130,000 (Park et al, 1986a), while the human receptor molecular weight has been estimated at 84,000 (DiPersio et al, 1988).

Although some of the properties of GM-CSF receptor have been deduced, the receptor has not hitherto been isolated or purified.

SUMMARY OF THE INVENTION

In accordance with the present invention, there is provided a purified nucleic acid selected from the group consisting of single stranded DNA, double stranded DNA, cDNA and RNA which each comprises a nucleotide sequence which encodes, or is complementary to a nucleotide sequence which encodes, human GM-CSF receptor or a derivative thereof capable of binding human GM-CSF, said GM-CSF receptor being defined as a glycosylated or unglycosylated proteinaceous molecule comprising, in its entirety, an extracellular domain, a transmembrane domain, and an intracytoplasmic tail and being capable of specifically binding radioactively labelled GM-CSF or derivatives thereof, said binding being competed for by unlabelled GM-CSF. The expression of this human GM-CSF gene provides a quantity of recombinant receptor heretofore unavailable, thereby permitting inter alia the development of receptor therapeutics, diagnostics, agonists and antagonists, and the like.

Accordingly, the first aspect of the present invention relates to recombinant or synthetic human GM-CSF receptor or a derivative thereof capable of binding human GM-CSF, essentially free of other proteins, said GM-CSF receptor being defined as a glycosylated or unglycosylated proteinaceous molecule comprising, in its entirety, an extracellular domain, a transmembrane domain, and an intracytoplasmic tail and being capable of specifically binding radioactively labelled GM-CSF or derivatives thereof, said binding being competed for by unlabelled GM-CSF. Derivatives included in this aspect of the invention are those including that portion of the receptor comprising its soluble (non-membrane associated) region that has the capacity to bind GM-CSF.

A second aspect of the present invention relates to antibodies that recognise human recombinant or synthetic GM-CSF receptor (and its derivatives) as hereinbefore defined, and which are useful in the detection and/or purification of these molecules.

The binding of GM-CSF to its cell-bound receptor may be modulated with agonists or antagonists of GM-CSF.

The receptors of the present invention may be used in a method of treatment of GM-CSF related diseases in a mammal, and in particular a human, by the administration to said mammal of an effective amount of recombinant or synthetic GM-CSF receptor or its derivative. One such method involves modulating the proliferation, differentiation or functional activation of GM-CSF-stimulation-sensitive cells in the mammal, which method comprises the administration to said mammal of an effective amount of recombinant or synthetic GM-CSF receptor of its derivative for a time and under conditions sufficient to reduce the amount of non-bound GM-CSF. An effective amount can readily be determined by routine experiment, as can the most effective and/or convenient route of administration, such as intravenous, intramuscular, subcutaneous, or oral. Slow release formulations may be advantageous for specific purposes. Preferably the mammal and the origin of the GM-CSF receptor are homologous, and even more preferably the homologous system is human.

A third aspect of the present invention, provides a method of generating information for use in the diagnosis of a GM-CSF related disease in a mammal, comprising the step of contacting a tissue or biological fluid or said mammal, in vitro, with a nucleic acid or recombinant DNA molecule, a recombinant cell containing such a molecule, a recombinant or synthetic GM-CSF receptor or an antibody or antigen-binding fragment thereof as hereinbefore defined, and detecting the thus-formed product. In this aspect, the invention also extends to kits for such diagnostic purposes. For example, diagnostic kits for assays utilizing radioimmunoassays, fluorescent immunoassay, or ELISA are specifically contemplated.

A fourth aspect of the present invention relates to the use of recombinant or synthetic human GM-CSF receptor or an antibody or antigen binding fragment thereof as hereinbefore defined in the manufacture of a medicament for the treatment of GM-CSF related diseases.

A fifth aspect of the invention relates to a GM-CSF-dependent haemopoietic cell line into which a low-affinity receptor for a heterologous GM-CSF has been cloned.

A sixth aspect of the invention relates to a method of screening a cDNA library for a cDNA fragment encoding GM-CSF receptor, comprising the steps of:
constructing a cDNA library, preparing cDNA fragments therefrom, transfecting said fragments into mammalian hosts cells, incubating said transfected cells with labelled GM-CSF, identifying populations of transfected cells binding said labelled GM-CSF, preparing clones of host cells transformed with cDNA fragments able to cause mammalian host cells to bind labelled GM-CSF, and isolating said clones.

BRIEF DESCRIPTION OF THE FIGURES

The present invention is further described by way of reference only to the following figures, in which:

Figure 1 shows saturation binding isotherms and Scatchard analysis of ¹²⁵I-hGM-CSF binding to

A. HL60 cells after incubation in 1.25% (w/v) DMSO for 5 days and

B. purified human placental membranes.

In (A), 5x10⁶ DMSO-treated HL60 cells per point were incubated with increasing concentrations of ¹²⁵I-hGM-CSF at 4°C for 3 hours and specific cell-associated radioactivity measured immediately after centrifugation or after 10 min incubation in 1 ml phosphate buffered saline (PBS) at 4°C. Scatchard analyses of these binding data are shown underneath before or after dissociation. The computer-derived high- and low-affinity binding components for undissociated cells are shown as solid lines. In (B), 40 µl of human placental membrane suspension, prepared as in Example 1, were incubated with increasing concentrations of ¹²⁵I-hGM-CSF at 20°C for 1 hour and specific binding subjected to Scatchard transformation is shown in the lower part of panel (B).

Figure 2 shows the binding specificity of hGM-CSF and h-IL-3 receptors on human placental membranes. Membranes (40 p1) were incubated in duplicate with ¹²⁵I-hGM-cSF (200,000 cpm in 110 µl HRF) with or without unlabelled hGM-CSF (500 ng) or h-IL-3 (100 ng) for 1 hour at 20°C. Similarly, 40 µl membranes were incubated with ¹²⁵I-h-IL-3 (220,000 cpm in 110 µl HRF) with or without the same amounts of unlabelled hGM-CSF or IL-3 as above. Total binding to each membrane preparatlon (mean ± range) is shown.

Figure 3 shows the detection and specificity of hGM-CSF receptors transfected to COS-7 cells.

(a) Photomicrograph of the cell autoradiograph showing the single clearly positive cell detected in 1.5x10⁶ COS-7 cells transfected with Pool 138 of the cDNA library (the COS cell covered by autoradiographic grains); magnification X20.

(b) Same as (a) but photographed under dark field illumination.

(c) Dark field illumination of a cell autoradiograph of COS-7 cells transfected with a pure hGM-CSF receptor clone (clone pGMR138) and incubated with 2nM ¹²⁵I-hGM-CSF; magnification X10.

(d) As for (c) except that cells were also incubated with 20 nM unlabelled hGM-CSF. Note dramatic reduction in autoradiographic grains.

(e) Specificity of the placental hGM-CSF receptor transfected to COS-7 cells. COS-7 cells transfected 48 hours earlier with the pure clone (pGMR138) were assayed in duplicate for their ability to bind ¹²⁵I-hGM-CSF in the absence or presence of 100 ng unlabelled hGM-CSF, h-IL-3, murine GM-CSF, human G-CSF or human IL-6 (mean ± range). Transfected cells (30,000 per point) were incubated for 1 hour at 20°C in HRF containing 20 mM EDTA and 100 µg/ml chondroitin sulphate and ¹²⁵I-hGM-CSF (70,000 cpm in 100 µl).

Figure 4 shows saturation and competition binding analysis of the placental hGM-CSF receptor transfected to COS-7 cells. COS-7 cells (33,000 per point) transfected 48 hours earlier with the pure clone (pGMR138) were incubated with increasing concentrations of ¹²⁵I-hGM-CSF (200,000 cpm) (Panel A) or a constant amount of ¹²⁵I-hGM-CSF and increasing concentrations of unlabelled hGM-CSF or h-IL-3 (Panel B) in a constant volume of 85 µl HRF/20 mM EDTA/100 µg/ml chondroitin sulphate at 20°C for 1.5 hours. In Panel A, total binding, non-specific binding and specific binding are shown with Scatchard transformation of the specific binding data shown underneath. In Panel B, total binding is shown in the upper panel, with Scatchard transformation of the specific binding data shown underneath.

Figure 5 shows analysis of chemical cross-linking of the hGM-CSF receptor on DMSO-treated HL60 cells and on transfected COS-7 cells, using SDS-polyacrylamide gel electrophoresis. In tracks A-D, 5x10⁶ HL60 cells treated for 9 days with DMSO were used per point while in tracks E-J, 7x10⁴ transfected COS-7 cells were used per point. In each case binding was for 3 hours at 4°C with ¹²⁵I-hGM-CSF at 2 nM. Tracks C and D were with or without 10 min dissociation in 1 ml PBS to remove low-affinity binding, and A and B are as for C and D except that 20 nM unlabelled hGM-CSF was included during the binding reaction. In each case 1 mM DSS was used for cross-linking (15 min on ice). Tracks E-J represent transfected COS cells allowed to bind ¹²⁵I-hGM-CSF and then cross-linked with E, 0mM; F, 0.01 mM; G, 0.05 mM; H, 0.1 mM, I, 0.5 mM and J, 1 mM DSS for 15 min on ice. Gel electrophoresis was on 10% (w/v) SDS gels (A-D) or 8% (w/v) SDS gels (E-J) and autoradiographs were exposed for 4 weeks (A-D) or 2 days (E-J). Molecular weight markers (Pharmacia and BioRad) are shown.

Figure 6(A) represents a restriction endonuclease cleavage map of the insert cDNAs of pGMR138 and pGMR29. Boxes represent open reading frames. The hatched and filled regions represent the signal sequence and transmembrane region of the GM-CSF-R coding region, respectively. The stippled box represents the upstream open reading frame.

Figure 6(B) shows the combined nucleotide sequence and deduced amino acid sequence of the insert cDNAs of pGMR138 and pGMR29. Numbers at the right margin indicate positions of nucleotides and numbers above the sequence refer to the amino acid sequence. The hatch marks indicate potential N-glycosylation sites (Asn-X-Ser/Thr). The overlined regions indicate the presumed signal peptide and transmembrane region, respectively. The sets of six asterisks identify possible poly(A) addition signals. The poly(A) tail in clone 138 was 61 nucleotides long.

Figure 7 is a hydrophobicity plot of the hGM-CSF receptor sequence according to the method of Hopp and Woods (1983) (span length=15). Note the two hydrophobic regions corresponding to the signal sequence and transmembrane domain, respectively.

Figure 8 is a photograph showing the detection of the human GM-CSF receptor transcript.

A. RNA from the following sources was probed for hGM-CSF receptor transcripts by Northern blot analysis as described in Example 1: CEM cells (Track 1); HepG2 cells (Track 2); HL60 cells (Tracks 3 and 4); HL60 cells cultured at 5x10⁵ cells/ml with 100ng/ml TPA for 3 days (Track 5). The positions of the 28S and 18S rRNA molecules are indicated. On other gels, RNA size standards (BRL) were also included. Exposure time of the autoradiograph: 5 days.

B. RNA from the following sources was probed for hGM-CSF receptor transcripts by PCR-based amplification of cDNA as described in Example 1: U937 cells (Track 1); AML193 cells (Track 2); HL60 cells (Tracks 5 and 6); CEM cells (Track 7); Raji cells (Track 8); Hep G2 cells (Track 9); HeLa cells (Track 10).

Tracks

3, 4 and 11 contained the following negative controls: "RNA blank" from no cells, carried through cDNA synthesis and PCR reactions; HL60 RNA not subjected to cDNA synthesis; and murine 160T cells, respectively.

Figure 9 depicts the alignment of the sequence of the hGM-CSF-receptor with other growth factor receptors. Amino acids are identified using the single letter abbreviations standard in the art.

A. Schematic of Alignment: Boxes represent coding regions. The hatched and solid boxes represent signal sequences and transmembrane regions, respectively. The positions of four conserved cysteine residues (C) and a conserved tryptophan residue (W) are indicated by vertical lines. The stippled box represents the "WS-WS" box. Sequences are aligned via the first conserved cysteine residue.

B. Details of Sequence Alignment: Reference positions (*) are marked above the sequence (i-vii) and correspond to the following amino acid numbers of each sequence (numbers in brackets refer to positions (i)-(vii), respectively): hGMR (126, 136, 165, 178, 236, 294, 331i, hIL6R (121, 132, 165, 176, 233, 290, 374), mEPOR (52, 62, 90, 106, 165, 219, NH), hIL2R (36, 46, 60, 74, 126, 182, NH) and rPRLR (31, 41, 70, 81, 146, 199, NH). NH: no homology; Cons: consensus sequence based on haemopoietic receptor sequences. Except in the consensus, dots indicate variable spacing and dashes indicate gaps introduced to align the sequences.

Figure 10 shows viral integration and transcripts in hGM-R-FD cells.

(a) DNA from FDC-P1 cells (track 1), hGM-R-

FD clones

1, 6, 8, 10, 11, 13, 21, 24, 33, 34, 49, 50, 52, 53, 54, 55, 56, 57 and 58 (tracks 2-20) and hGM-R-FD clones 21, 21.13, 21.15, 21.17, 21.21, 21.22 and 21.23 (tracks 21-27) was digested with Pst I and probed for hGM-R sequences as described in Example 1. Note the common 2.5 Kbp fragment derived from within the hGM-R viral construct

(b) Total cytoplasmic RNA from FDC-P1 cells (track 1) and hGM-R-FD clones 21.13, 21.15, 21.17, 21.21, 21.22, 21.23, 21.7, 21.8, 21.10 and 21.11 (tracks 2-11) was probed for viral hGM-R transcripts as described in Example 1. The major species is 5.5 Kb in length and corresponds to the full-length unspliced viral transcript.

Figure 11 illustrates the responsiveness of cells from cloned hGMR-FD cell lines to proliferative stimulation by recombinant m- or hGM-CSF. Lines maintained in human GM-CSF exhibit a similar content of clonogenic cells with both stimuli (eg. clone 54), whereas in lines maintained in m+hGM-CSF (eg. clone 21) murine-responsive clonogenic cells are more frequent than human-responsive cells. For both types of cell line, clonogenic cells are less responsive to human than to murine GM-CSF.

Figure 12 shows saturation binding analysis and Scatchard transformation of ¹²⁵-hGM-CSF binding to hGM-R-FD cell clones.

A, B : clone 21 maintained in h+mGM-CSF, 2x10⁶ cells per point;

C, D : clone 50 maintained in hGM-CSF only; 0.8x10⁶ cells per point.

A and C are specific binding curves with increasing amounts of ¹²⁵I-hGM-CSF added and B and D are the Scatchard transformations. The Scatchard transformations gave K_D=4nM, 4,000 receptors/cell for clone 21 and K_D=6nM, 20,000 receptors/cell for clone 50.

Figure 13 illustrates the internalization of ¹²⁵I-hGM-CSF bound to hGM-R-FD clone 33 cells (upper curve) or hGM-R-FD clone 53 cells (lower curve) at 37°C. The former cells had been maintained in a mixture of murine and human GM-CSF while the latter had been maintained in hGM-CSF only. The curves show the variation of cell surface-associated and internalized radioactivity with time, after addition of ¹²⁵I-hGM-CSF, determined as described (Nicola et al, 1988); the lines through the experimental points (means of duplicate tubes) were fitted by computer as described by Nicola et al (1988). For clone 33, 1.9x10⁶ cells were used per point and the ¹²⁵I-hGM-CSF concentration was 13nM, and for clone 53, 1.8x10⁶ cells were used per point and the ¹²⁵I-hGM-CSF concentration was 13nM.

DETAILED DESCRIPTION OF THE INVENTION.

The following abbreviations are used herein:

h,mGM-CSF: human or murine granulocyte-macrophage colony-stimulating factor
G-CSF: granulocyte colony-stimulating factor
multi-CSF: multipotential colony stimulating factor
IL-2: interleukin-2
HRF: Hepes buffered (10 mM, pH 7.4) RPMI medium containing 10% v/v foetal calf serum
PBS: phosphate buffered saline
EDTA: ethylene diamine tetraacetic acid
FCS: foetal calf serum
IL-3: interleukin-3
IL-6: interleukin-6
EPO: erythropoietin
PRL: prolactin
h-: human
m-: murine
r-: rat
LIF: leukaemia inhibitory factor
PCR: polymerase chain reaction
ORF: open reading frane
h-GM-R: human receptor for GM-CSF
h-GM-R-FD: FDC-P1 cells transfected with hGM-R
IL2R: IL-2 receptor (β chain)
IL6R: IL-6 receptor
GMR, GM-CSF-R: GM-CSF receptor
K_D: equilibrium dissociation constant
EPOR: EPO receptor
PRLR: PRL receptor
SDS: sodium dodecyl sulphate
SSC: standard saline citrate
dNTP: deoxynucleoside triphosphate
DSS: disuccinimidyl suberate

By "GM-CSF receptor" is meant a glycosylated or unglycosylated proteinaceous (i.e. amino acid containing) molecule comprising, in its entirety, an extracellular domain, a transmembrane domain and an intracytoplasmic tail. The receptor molecule is capable of specifically binding radioactively labelled GM-CSF or derivatives thereof, said binding being defined inter alia as being competed for by unlabelled GM-CSF.

By "recombinant GM-CSF receptor" is meant a glycosylated or unglycosylated polypeptide molecule, with or without other associated molecules (eg lipids), produced by recombinant means such as by the ligation of a cDNA molecule encoding the receptor or its derivative into an appropriate expression vector in the correct reading frame relative to a promoter, introducing the resultant recombinant expression vector into a suitable host and growing said host under conditions appropriate for the expression and, if necessary, transportation of the recombinant receptor or its derivative from said host, and then purifying the recombinant receptor or derivative.

By "GM-CSF-stimulation-sensitive cell" is meant a cell carrying a receptor to which GM-CSF can bind thereby causing the stimulation of proliferation or functional activation of that cell, as hereinbefore described. The term "bind" in relation to GM-CSF and its receptor is used in its broadest sense, and means any association between GM-CSF and its receptor, and particularly in relation to cell-bound receptor, sufficient to induce the stimulation of proliferation or functional activation of the cell on which the receptor is located. "Non-bound GM-CSF" generally means circulating GM-CSF.

By "substantial amino acid homology" is meant molecules having a sequence homology of approximately 75% or more, preferably greater than or equal to 85% and even more preferably greater than or equal to 90-95%.

"Cancer" is used in its broadest sense, and includes cancers, tumours and leukaemias en masse or as individual cells. "Aberrations" of bound GM-CSF receptor or its nucleotide sequence is used to mean any alteration in amino acid and/or nucleotide sequence detectable by analysis of the respective homologues as determined, for example, by hybridization studies. Accordingly, a cancer cell may contain an altered GM-CSF receptor detectable as being so altered by the use of the GM-CSF receptor or its encoding nucleotide sequence contemplated by the present invention.

General techniques for cloning and expressing cDNA are well known in the art, and can be found, for example, in Maniatis et al (1982). For the purposes of exemplification only, the present invention is described with reference to the cDNA encoding a hGM-CSF receptor expressed in COS-7 cells employing the COS cell expression vector nH3M (Aruffo and Seed, 1987). This is done with the understanding that the present invention includes within its scope cDNA encoding hGM-CSF receptor expressed in any vector and/or host. For example, it would be a matter of routine experimentation for a person skilled in the art to place the subject cDNA into a prokaryotic expression vector and express same in bacteria, such as Escherichia coli, Bacillus sp. or Pseudomonas species. By routine manipulation, the cDNA can be expressed in eukaryotic cells such as yeasts, fungi, insect cell lines, mammalian cell lines other than COS cells, or plant cells. By known methods, the cDNA can also be transferred into germ line or somatic cells to form transgenic animals.

The present invention extends to single or double stranded DNA, cDNA or mRNA wherein at least one nucleotide strand thereof encodes, or is complementary to, a nucleotide strand which encodes human GM-CSF receptor or its derivatives, and to any vector, expression or otherwise, containing same, including viral vectors.

In a preferred embodiment, the invention provides a recombinant DNA molecule comprising a gene encoding the human low-affinity GM-CSF receptor, or a homologue thereof which encodes a polypeptide having a least 75% sequence identity with said receptor and which retains at least one-tenth of the relative binding affinity of the native human low-affinity GM-CSF receptor for human GM-CSF, said gene being operably linked to a promoter. More preferab said polypeptide comprises an extracellular domain, a transmembrane domain, and an intracellular domain.

Although the sequence presented in Figure 6B refers specifically to the low-affinity GM-CSF receptor, the evidence presented herein indicates that the low affinity GM-CSF receptor itself forms a component of the high-affinity GM-CSF receptor, which may be a multimer of the low-affinity GM-CSF receptor. The high-affinity GM-CSF receptor is therefore specifically included in the scope of the present invention.

In accordance with a particularly preferred embodiment of the present invention, there is provided the GM-CSF receptor molecule having the amino acid sequence shown in Figure 6B. This molecule is a 400 amino acid polypeptide with a molecular weight of approximately 45,000, having a single hydrophobic transmembrane domain, a glycosylated extracellular domain and a short (54 amino acid) intracytoplasmic tail. The ligand-binding domain (23-319) contains eleven cysteine residues. The subject receptor does not contain a tyrosine kinase domain, and does not show homology with members of the immunoglobulin gene superfamily, but does share sequence homology with receptors of other haemopoietic growth factors such as human IL-6, erythropoietin and IL-2 (β-chain). This GM-CSF receptor shows a single class of binding affinity (K_D=2-8nM), with specificity for GM-CSF but not IL-3.

It is well within the capacity of persons skilled in the art to construct a cDNA sequence encoding the receptor molecule as shown in Figure 6B using methods which are well established and materials which are readily available, as described herein.

Given the aforementioned amino acid sequence and biochemical properties, the present invention includes within its scope a synthetic GM-CSF receptor molecule prepared by the chemical addition of amino acids by established techniques and in a sequence as established herein. The invention further includes recombinant or synthetic derivatives of GM-CSF receptor carrying single or multiple amino acid substitutions, deletions and/or additions to any or all of the aforementioned regions or domains of the receptor molecules. Such derivatives may be functional (i.e. biological) equivalents with respect to their ability to bind to GM-CSF or its derivatives, and/or may exhibit substantial amino acid homology to the aforementioned GM-CSF receptor amino acid sequence. One preferred derivative of GM-CSF receptor comprises all or part of the extracellular domain (soluble portion). Derivatives also include any or all of the receptor molecule in glycosylated or unglycosylated form. Depending upon the expression system and host employed, for example, the recombinant receptor may or may not be glycosylated. Both glycosylated and unglycosylated forms of the recombinant or synthetic GM-CSF receptor or its derivatives are within the scope of the present invention. Functionally active derivatives or equivalents of the GM-CSF receptor can readily be identified using the methods described herein.

The invention is also directed inter alia to the cDNA encoding a GM-CSF receptor as hereinbefore defined. The cDNA comprises a nucleotide sequence shown in Figure 6B. The scope of the present invention includes cDNA derivatives carrying single or multiple nucleotide substitutions, deletions and/or additions relative to the aforementioned cDNA sequence. Such derivatives may encode the entire GM-CSF receptor molecule or derivatives thereof, such as a GM-CSF receptor carrying single or multiple amino acid substitutions, deletions and/or additions. The present invention also encompasses a cDNA carrying a nucleotide sequence substantially homologous to the subject nucleotide sequence, i.e. carrying at least 75% homology, preferably 80-85% homology and even more preferably greater than 90-95% homology. Methods for producing such derivatives, for example by site-directed mutagenesis, random mutagenesis, or enzymatic cleavage and/or ligation of the nucleic acid, are well known in the art, as are methods for determining whether the thus-modified nucleic acid has significant homology to the subject sequence, for example by hybridization.

The present invention further includes molecules such as polypeptides fused to the GM-CSF receptor or its derivatives, or nucleotide sequences contiguous to GM-CSF receptor-encoding sequences. For example, and not by way of limitation, it may be desirable to produce a fusion protein comprising GM-CSF receptor or its derivative and an amino acid sequence from another polypeptide or protein, examples of the latter, especially prokaryotic systems, being enzymes such as β-galactosidase, phosphatase, urease and the like. Most fusion proteins are formed by the expression of a recombinant gene in which two coding sequences have been joined together such that their reading frames are in phase. Alternatively, polypeptides can be linked in vitro by chemical means. All such fusion or hybrid derivatives of GM-CSF receptor or the respective encoding nucleotide sequences are encompassed by the present invention.

The present invention, therefore, provides recombinant and synthetic GM-CSF receptor and derivatives thereof in sufficient quantity for the development of, for example, receptor therapeutics and diagnostics.

Accordingly, another aspect of the invention contemplates a method for modulating the proliferation or functional activation of GM-CSF-stimulation-sensitive cells in a mammal which method comprises the administration to said mammal of an effective amount of GM-CSF receptor or its derivative for a time and under conditions sufficient to reduce the amount of non-bound GM-CSF. The subject method is predicated on the binding of GM-CSF circulating in the body to soluble receptor before the lymphokine can bind to the cell-bound receptor, thereby reducing the amount of GM-CSF available for binding to GM-CSF-stimulation-sensitive cells.

The invention extends to the use of the entire recombinant or synthetic GM-CSF receptor or its derivative in the modulation of proliferation or functional activation of GM-CSF-stimulation-sensitive cells, but preferably the soluble or extracellular domain is used.

This method is particularly useful in treating disease states such as myeloid leukaemias where the leukaemic cells express GM-CSF receptors (Young and Griffin, 1986). Many types of myeloid leukaemia cells, for example, cannot undergo proliferation in vitro unless exogenous colony stimulating factors are available (Metcalf, 1984). Accordingly, recombinant or synthetic GM-CSF receptor or its derivatives administered in vivo will bind to circulating GM-CSF and thus compete for binding of the latter to cell bound GM-CSF receptor. The subject method will also assist as therapeutic agents in abrogating the toxic effects of over-production of GM-CSF noted in animal model systems (Lang et al, 1987; Johnson et al, 1989).

The present invention also extends to antibodies, and particularly monoclonal antibodies, to recombinant or synthetic GM-CSF receptor or its derivatives. Such antibodies are particularly useful in purifying GM-CSF receptors and for quantitating the same. The subject invention also extends to antibodies (monoclonal or polyclonal) against said first antibodies for the purposes of assaying plasma, serum or body fluids, cell surfaces or cell extracts for the presence of GM-CSF receptors. One or other of the aforementioned antibodies may be labelled with a reporter molecule for use, for example, in sandwich assays. Methods for production and screening of polyclonal and monoclonal antibodies, optionally including the use of adjuvants, methods for labelling such antibodies with radioactive, fluorescent, or chemical labels, and methods of immunoassay, such as radioimmunoassay, fluorescent immunoassay, and enzyme-linked immunoassay (ELISA), and coupling of antibodies to solid supports to form immunoadsorbents, are routine in the art.

Furthermore, the recombinant or synthetic GM-CSF receptors can be used in developing agonists and antagonists to augment or reduce binding of GM-CSF to its cell-bound receptors. For example, a transfected cell line bearing recombinant GM-CSF receptors may be used to screen compounds for such activities. The present invention extends to the use of such agonists or antagonists in the treatment of disease states caused by excessive or insufficient proliferation or functional activation of GM-CSF-stimulation-sensitive cells. Possible agonists or antagonists include naturally occurring or synthetic fragments of GM-CSF, as well as other naturally occurring or synthetic chemical compounds, which may be screened by the aforesaid method, using binding of labelled GM-CSF to the cells as a marker.

The present invention also extends to the use of GM-CSF or its derivatives in the manufacture of a medicament for the treatment of GM-CSF-related diseases. Such diseases include cancers, tumors and leukaemias caused by or associated with GM-CSF-stimulation-sensitive cells.

Another aspect of the present invention extends to the use of recombinant or synthetic GM-CSF receptor or derivatives thereof, and particularly to nucleotide sequences encoding same, in the diagnosis of cancers composed of, or associated with, GM-CSF-stimulation-sensitive cells by detecting cell-bound GM-CSF receptors or aberrations thereof, or nucleotide sequences encoding same, in said cancer cells.

In accordance with the present invention, the functional, low-affinity hGM-CSF receptor has been cloned from placenta, and it has been shown that it recognizes only GM-CSF and not interleukin-3. This receptor, when transfected into COS cells, shows nearly identical low affinity to, and the same specificity as the receptor on placental membranes. Moreover, like the low affinity hGM-CSF receptor on haemopoietic cells, the transfected receptor is characterized by a rapid ligand dissociation rate (T_1/2=5 min) and poor internalization (approximately 10-20% after 2 hours at 37°C).

Furthermore, as shown in Figure 9, the positions of four cysteine residues are approximately conserved in the receptors for GM-CSF, IL-6, erythropoietin and IL-2 (β-chain) respectively, and the context in which three of these residues are found is identical between the four receptor types. These four cysteine residues, which might form pairs of intra-chain disulphide bonds, do not coincide with either of the two cysteine residues of the IL-6 receptor associated with the putative immunoglobulin domain structure (Yamasaki et al, 1988).

Trp236 is conserved, and in three of the receptors is adjacent to an Arg residue. A further homology between all four receptor sequences is found just N-terminal of the transmembrane domain (Figure 9). The consensus sequence starting at position 294 of the GM-CSF receptor (the "WS-WS" Dox) is found in all four receptors. The position of this sequence is close to the transmembrane domain in three of the receptors (for GM-CSF, erythropoietin and IL-2 (β-chain)), but is further away in the IL-6 receptor.

A string based on the "WS-WS" box (VXXRXX_(6-11)WSXWS) was used to search the most up-to-date database (Protein Research Foundation, Japan; April 1989), and a region homologous to this string was found in the rat prolactin receptor (rPRL receptor; Boutin et al, 1988) just N-terminal to its transmembrane domain (Figure 9). Strikingly, all four of the extracellular cysteine residues ln the rPRL receptor, as well as a Lys-Trp doublet, are found in approximately the same relative positions as those of the four haemopoietic receptors (Figure 9A). The rPRL receptor also has a transmembrane cysteine residue (Boutin et al, 1988), but the transmembrane sequence is not homologous to the hGM-CSF receptor. By contrast, the "WS-WS" box is not found in a close relative of the prolactin receptor, the growth hormone receptor (Leung et al, 1988), which in other regions shares 75-100% sequence similarity with the prolactin receptor (Boutin et al, 1988). The hGM-CSF receptor therefore appears to be a member of a new subset of growth and differentiation factor receptors, defined by the set of five receptors described above (i.e. hGM-CSF, hIL-6, murine EPO, hIL-2 (β-chain) and rPRL receptors).

In the GM-CSF receptor mRNA, there is a short 22 codon ORF prior to the long ORF encoding the GM-CSF receptor. Interestingly, such short ORFs are also found 5' of the main receptor coding region in the DNA sequences for the human IL-6 receptor (Yamasaki et al, 1988), the murine IL-1 receptor (Sims et al, 1988) and the human IL-2 receptor α and β-chains (Nikaido et al, 1984; Hatekeyama et al, 1989) and might act, if translated, to depress the translation of the main receptor coding regions. Such a mechanism might partly explain the low levels of expression of these receptors in normal cell types.
The present invention is further described by way of reference only to the following non-limiting examples:

EXAMPLE 1 MATERIALS AND METHODS

The following materials or methods were employed in subsequent examples. Biological starting materials referred to herein are known in the art.

A cDNA library derived from poly-A+ selected human placental RNA, constructed in the COS cell expression vector nH3M (Aruffo and Seed, 1987) and consisting of approximately 5x10⁶ independent clones, was provided by Dr. B. Seed (Massachusetts General Hospital, Boston, USA). Approximately 10⁷ clones were produced by transforming MC1061/p3 cells, sorting them into 500 pools of approximately 2x10⁴ clones, and preparing glycerol stocks. Miniprep DNA from each was transfected into COS-7 cells by electroporation. Briefly, 1.5x10⁶ COS-7 cells in 180µl phosphate-buffered saline pH7.3 (PBS) were mixed with 20p1 miniprep DNA (3µg) and chilled on ice for 5 min. Cells were electroporated in 0.4cm gap cuvettes at 300V and 125pFD (t_c=8.2-10.5msec), returned to ice for 5 min and finally cultured in 2ml Dulbecco's modified Eagle's medium (DME) containing 10% (v/v) foetal calf serum (FCS) in glass slide-based flaskettes (Lab-Tek, Nunc Inc., Naperville, USA). These conditions led to a 15-20% transfection frequency in surviving COS-7 cells as assessed by control transfections of ICAM/CDM8 (Simmons et al, 1987) labelled with a radioiodinated anti-ICAM monoclonal antibody W-CAM-1 (Boyd et al, 1988). After 48 hours, the medium was removed and the transfected monolayers assessed for binding of radioiodinated human GM-CSF (4-8x10⁵ cpm of ¹²⁵I-GM-CSF (1-2nM) in 1ml Hepes buffered RPM1 medium pH7.2/10% FCS) for 60 min at 20°C. The monolayers were washed twice in medium, fixed in 2.5% (w/v) glutaraldehyde/PBS and dipped in 1% (w/v) gelatine as described (Nicola and Metcalf, 1985). The slides were dipped in Kodak NTB2 photographic emulsion at 42°C and exposed in the dark for 48 hours at 4°C in light-proof boxes containing dessicant. Slides were developed for 3 min in Kodak D19 developer (40g/500ml water), and rinsed in water and fixed for 3 min in Agfa G433C fixer before staining in 10% filtered Giemsa stain in water. Slides were screened at 10-20x magnification and two positive pools (#29 and #138) were selected. Corresponding glycerol stocks of E.coli transformants were partitioned into smaller pools until single cDNA clones able to cause COS-7 cells to bind ¹²⁵I-GM-CSF were obtained.

Sequencing Strategy

The inserts and various internal fragments of clones 29 and 138 were subcloned into M13 vectors and sequenced by the dideoxy chain termination method (Sanger et al, 1977) using modified T7 polymerase (Tabor and Richardson, 1987; Sequenase, USB) and primers internal to subcloned segments. Ambiguities were resolved using dITP. Primers corresponding to some of these subclones were used to extend the sequence to adjacent segments. Both strands of clone 138 were sequenced in their entirety (average gel characters per contig character=4.87). All subcloning boundaries were resequenced on full-length clones. The mRNA-synonymous strand of clone 29 was completely sequenced, and ambiguities resolved using oligonucleotides on the opposite strand.

Analysis of RNA and DNA (Northern and Southern Blotting)

Cytoplasmic polyadenylated RNA (approximately 1.5µg), prepared essentially as described by Gough (1988), was fractionated on 1% (w/v) agarose gels containing 20mM morpholinopropane sulphonic acid, 5mM sodium acetate, lmM EDTA (pH 7.0), plus 6% (v/v) formaldehyde, and transferred to nitrocellulose. Prior to hybridization, filters containing RNA were soaked in 2xSSC containing 0.2% (w/v) Ficoll, 0.2% (w/v) polyvinylpyrollidone, 0.2% (w/v) bovine serum albumin, 2mM sodium pyrophosphate, 1mM ATP, 30 50µg/ml denatured salmon sperm DNA and 50µg/ml E. coli tRNA at 67°C for several hours. Hybridization was performed in the same buffer plus 0.1% (w/v) SDS at 67°C. The hybridization probe was the gel-purified 1300bp Xho1-EcoRI. fragment spanning the 5' end of cDNA clone pGMR138, radiolabelled to a specific activity of approximately 10⁹ cpm/µg by random priming (Feinberg and Vogelstein, 1983) and included in the hybridization at approximately 5x10⁷ cpm/ml. Filters were washed extensively in 2xSSC, 0.1% (w/v) SDS at 67°C and finally in 0.2xSSC at 67°C prior to autoradiography.

Ten µg aliquots of high molecular weight genomic DNA were digested with PstI, electrophoresed in 0.8% agarose gels and transferred to nitrocellulose. Conditions for hybridization and washing were as described above for RNA analysis. The hybridization probe was the gel-purified 786bp Kpn-EcoRI fragment spanning the 3' end of the hGM-CSF receptor coding region, radiolabelled to a specific activity of approximately 2-4x10⁸ cpm/µg by nick-translation and included in the hybridizations at approximately 2x10⁷ cpm/ml.

RNA detection using the polymerase chain reaction

RNA (approx 1µg) was subjected to first strand cDNA synthesis in a 20µl reaction containing 50mM Tris-Cl (pH 8.3 at 42°C), 20mM KCl, 10mM MgCl₂, 5mM dithiothreitol, lmM of each dNTP, 20µg/ml oligo-dT₁₅ and 20 units AMV reverse transcriptase (Boehringer Mannheim) for 40 minutes at 42°C. After completion of first-strand synthesis, the reaction was diluted to 100 µl with distilled water and 5 pl used for each PCR (polymerase chain reaction). Polymerase chain reactions contained 200 µM of each dNTP, 1 µM of each specific primer, buffer as supplied in the GeneAmp kit (Cetus Corp., USA) and 1.25 units Taq polymerase in a volume of 50 µl. The primers used for PCR were 5'-CTTCTCTCTAGACCAGCA (position 131-147) and 5'-ACATGGGTTCCTGAGTC (position 676-660) defining a 530bp fragment. The PCR reaction conditions were: 2 min at 94°C; 2 min at 65°C; 3 min at 72°C for 25 cycles in a Perkin-Elmer-Cetus DNA thermal cycler. A portion of the PCR reaction was electrophoresed through a 1.2% (w/v) agarose gel and transferred to nitrocellulose. Filters were prehybridized, hybridized and washed as described above. The hybridization probe was the gel-purified 1.9kbp cDNA insert of pGMR138, radiolabelled to a specific activity of approximately 10⁹ cpm/µg by random priming and included in the hybridization at approximately 2x10⁷ cpm/ml.

Radioligands

Purified, recombinant human or mouse GM-CSF produced in non-glycosylated form in E. coli (Delamarter et al, 1985) was radioiodinated by the modified iodine monochloride method of Nicola et al (1988); briefly, 2 pg of protein (2 pl) was incubated with lmCi Na¹²⁵I (New England Nuclear, Dreieich, W. Germany) in 40 pl of 0.2M Na phosphate buffer pH 7.2 containing 0.2% (w/v) Tween 20. Iodine monochloride (0.03mM in 2M NaCl) was added in two lots of 3 µl and 6 µl while vortex-mixing the solution. The reaction mixture was passed through a column of Sephadex G-25M (Pharmacia, Uppsala, Sweden) to separate macromolecular radioactivity from free iodine (Hilton et al, 1989). ¹²⁵I-hGM-CSF was 100% bindable (Calvo et al, 1983) and displayed a specific radioactivity of 20,000-40,000 cpm/ng by the self-displacement analysis of Calvo et al(1983). ¹²⁵I-mGM-CSF had specific radioactivity 120,000 cpm/ng and bindability 40-50%. Unlabelled and labelled (specific radioactivity 40,000 cpm/ng) human IL-3 was purchased from Amersham (Buckinghamshire, England).

Binding Experiments

HL60 cells grown for 5 days in DME medium/10% FCS containing 1.25% (w/v) dimethyl sulphoxide were resuspended at 5x10⁶ cells/50 µl in Hepes (10mM, pH 7.2) buffered RPMI medium (HR) containing 10% (v/v) foetal calf serum (HRF). Fifty µl aliquots of cells were incubated with increasing concentrations of ¹²⁵I-hGM-CSF (0-2nM) with or without unlabelled hGM-CSF (0.3µm) at 4°C for 4 hours. Cell suspensions were then layered over 180 µl chilled foetal calf serum and centrifuged for 5 min at 700 g in small plastic centrifuge tubes and the cell pellet removed by cutting the tube with a scalpel blade. Cell-associated and free radioactivity were separately determined by counting duplicate tubes in a gamma-counter. Transfected COS-7 cells were harvested 48-72 hours after transfection by removing the supernatant and incubating the adherent cells in 40mM EDTA in HR containing 200 µg/ml chondroitin sulphate, and incubating at 37°C for a further 40 min. (Padmanabhan et al, 1988). The detached and disaggregated cells were centrifuged at 700 g for 5min and resuspended in HRF with or without 20 mM EDTA and 100 µg/ml chondroitin sulphate. Saturation binding isotherms or competition experiments were performed as for HL60 cells. Human placental membranes were prepared from fresh term placentas essentially as described by Yeung et al (1987), with 6g of placenta yielding 4 ml of membrane suspension. For each binding point, 40 µl of membrane suspension was mixed with 40 µl HRF and increasing concentrations of ¹²⁵I-hGM-CSF with or without excess unlabelled hGM-CSF (0.0 µM). After 1 hour incubation at 20°C the membranes were centrifuged at 30,000g for 5 minutes, the supernatant removed with a fine Pasteur pipette and membrane pellets and supernatants counted separately in a gamma-counter.

Saturation binding of ¹²⁵I-hGM-CSF to hGM-R-FD cells and Scatchard transformations were performed as described above, using an incubation time of lhr at 20°C. Cross-modulation of hGM-CSF receptors by mGM-CSF and of multi-CSF or mGM-CSF receptors by hGM-CSF was performed as described by Walker et al (1985), with preincubation time of 30 min at 37°C followed by binding times of 3hr at 0°C. Receptor internalization studies were performed at the indicated concentrations of ¹²⁵I-hGM-CSF, and the data were analyzed by curve fitting of the experimental points (Nicola et al, 1988).

Cross-linking Experiments

Binding of ¹²⁵I-hGM-CSF to cells in solution was performed at 4°C as described above, and the cell pellets resuspended in 1 ml of ice-cold Na phosphate buffered (20mM, pH 7.2) saline (0.15M). Disuccinimidyl suberate (Sigma, Missouri, USA) in anhydrous acetonitrile (10 µl) was immediately added to give a final concentration of 0-lmM, and the cells incubated for 15 minutes on ice before centrifuging the cell pellet at 13,000 g for 1 minute. The cell pellet was treated with DNAase in the presence of protease inhibitors and prepared for sodium dodecyl sulphate polyacrylamide gel electrophoresis as described (Nicola and Peterson, 1986).

Analysis of Binding Data

Specific binding was determined as the difference between binding in the absence or presence of excess unlabelled hGM-CSF. Specifically bound cpm were converted to molar concentrations using the specific radioactivity of ¹²⁵I-GM-CSF determined by self-displacement analysis. Curve-fitting of the binding data was performed using the LIGAND program of Munson and Rodbard (1980) before conversion to the Scatchard transformation. Two binding site fits were used only if the fit to the data was significantly improved (p<0.05) over the one binding site fit.

Production and Selection of hGM-R Retrovirus

A 1.7kbp Xhol fragment containing the insert cDNA of pGMR29, prepared as described herein, was inserted into the Xhol site of pJZen2(SVNeo), a derivative of the retroviral vector pMPZen (Johnson et al, 1989) in which a multiple cloning site has replaced the single Xhol site of pMPZen, and a 1.7kbp neomycin-resistance expression cassette (the BamHI-EcoRI fragment of pDol (Korman et al, 1987), blunt-ended and ligated to ClaI linkers) was inserted into the ClaI site (J. Chang, O. Bernard and K. Klingler, unpublished data). ψ2 packaging cells (Mann et al, 1983) were electroporated with pJZen2(SVNeo)-hGM-R DNA as previously described (Johnson et al, 1989), and transfectants were selected 2 days later using 400pg/ml of antibiotic G418 (Geneticin, Sigma). G418-resistant 2 clones were selected for high surface expression of hGM-R by bindlng of ¹²⁵I-hGM-CSF. Retroviral titres of receptor-positive ψ2 clones were tested by polybrene-mediated infection of NIH3T3 fibroblasts (Cepko et al, 1984). The clone selected for further work (ψ2-GMR) had a titre of 1.2x10⁴ viral particles/ml.

Derivation of Infected FDC-P1 Cell Lines

Adherent ψ2-GMR cells (3x10⁶ per 75cm² flask) were irradiated (35 Gy), and co-cultured with 10⁶ FDC-P1 cells (Dexter et al, 1980) in 20 ml Dulbecco's Modified Eagle's Medium (DMEM) wlth 10% foetal calf serum (FCS) and 10% pokeweed-mitogen stimulated spleen cell conditioned medium. Washed supernatant cells from 48 hour cocultures were cultured in agar-medium at a density of 300 cells/ml with either 10³ U/ml of mGM-CSF, 6x10³ U/ml of hGM-CSF, or a combination of both. After 3 days' incubation, clones developing in mGM-CSF had achieved 50-100 cells in size. In contrast, fewer clones had developed in cultures stimulated by hGM-CSF; these were dispersed in morphology, and most contained only 10-30 cells. Individual clones growing in cultures stimulated by hGM-CSF were removed using a micropipette, and cloned cell lines established and maintained in 1 ml cultures of DMEM with 20% FCS containing either 6x10⁵ U/ml hGM-CSF (12 lines) or 6x10³ U/ml hGM-CSF plus 10³ U/ml mGM-CSF (36 lines). Subcloning of individual cell lines was performed by growing colonies in agar medium cultures of 200 cells/ml then removing individual colonies after 7 days of incubation and continuing culture of these colonies in suspension.

Agar Cultures

These were performed in 35mm plastic Petri dishes (Nunc, Adelaide) using lml of agar medium (DMEM with final concentrations of 20% FCS and 0.3% agar) (Metcalf, 1984) and 300 cultured cells.

The stimuli used for colony formation were purified recombinant mGM-CSF (specific activity 3x10⁸ U/mg protein) or purified recombinant hGM-CSF (specific activity 10⁸ U/mg protein) produced as nonglycosylated derivatives in E.coli. These were included as 0.lml volumes during preparation of the agar cultures, and serial twofold dilutions were performed using 5% FCS in 0.95% saline. Mean cell numbers in 7-day colonies were determined by pooling 30-50 sequential colonies.

EXAMPLE 2 Detection of High and Low Affinity GM-CSF Receptors

In accordance with the methods described herein, both high and low affinity GM-CSF receptors were detected on human bone marrow cells, primary human myeloid leukaemic cells and the human promyelocytic leukaemic cell line, HL-60 (Fig. 1). Fig 1A shows the saturation binding isotherm at 4°C for ¹²⁵I-hGM-CSF binding to HL-60 cells that had been induced to differentiate for 5 days in 1.25% (w/v) dimethylsulphoxide (DMSO). Transformation of this data by the method of Scatchard (1949) after curve-fitting using the LIGAND program of Munson and Rodbard (1980) indicated that two, rather than one, binding sites were required to fit the data (p<0.05), with the high affinity receptors (40/cell) having a K_D of 46pM and the low affinity receptors (130/cell) having a K_D of 2.9nM. After allowing the same cells to dissociate bound GM-CSF for 10 minutes at 4°C and determining the remaining bound ¹²⁵I-hGM-CSF on the cells, only high affinity binding sites could be detected (Fig 1A). This indicates that the low affinity receptors display a more rapid rate of ligand dissociation than do high affinity receptors. This may explain the failure of other investigators (Gasson et al, 1988; Park et al, 1986b; Kelleher et al, 1988) to detect the low affinity receptors, since the subject binding procedure involves a rapid, single-step separation of cell-bound from free ligand. These authors described a single affinity class of GM-CSF receptors, having K_D of 20-600 pM, on human haemopoietic cells.

Specific binding of hGM-CSF to purified human placental cell membranes was also detected (Fig 1B), but this binding was to a single class of low-affinity receptor (3x10⁹ receptors/mg placenta, K_D = 4.6M). The human placental GM-CSF receptor appeared to recognize only hGM-CSF and not h-IL-3 (Fig 2). Moreover, no hIL-3 receptors were detected on human placental membranes (Fig 2).

EXAMPLE 3 Cloning and Expression of Human GM-CSF Receptor

In order to clone the GM-CSF receptor, an expression screening strategy was employed in simian COS cells. The low abundance of the GM-CSF receptor in all cell sources so far examined necessitated the screening of large cDNA libraries and consequent sensitive detection of positive transfected cells. The detection procedure we have adopted involves binding ¹²⁵I-hGM-CSF to transfected COS cells grown on glass microscope slides, followed by direct autoradiography. This screening approach has two major advantages over previous receptor cloning procedures which used radioiodinated ligand as their detection tool (Sims et al, 1988; D'Andrea et al, 1989). First, this procedure is extremely sensitive, since single clones could be detected from pools of 2x10⁴ clones (c.f. 1 in 10³, D'Andrea et al, 1989; 1 in 350, Sims et al, 1988). Secondly, the procedure allows easy identification of artefacts due to non-specific binding, which would otherwise have made identification of a single GM-CSF-receptor positive cell on a slide containing approximately 10⁶ negative COS cells impossible.

A human placental cDNA library of approximately 5x10⁶ independent recombinants in a COS cell expression vector (Aruffo and Seed, 1987) was sub-fractionated into 500 pools of approximately 2x10⁴ clones each, and DNA from each pool was separately transfected into 1.5x10⁶ COS cells by electroporation. The cells were cultured for 48 hours on glass slides, incubated with a relatively high concentration of ¹²⁵I-hGM-CSF (approximately 2nM), fixed and then dipped in liquid photographic emulsion. The slides were developed and individually examined microscopically; positive cells were identified by the presence of autoradiographic grains. Of the first 250 pools screened, two gave rise to 1 or 2 positive cells (pools 29 and 138) (Figure 3). One of these pools (138) was partitioned into smaller pools (20 pools of 1500 recombinants then 60 pools of 80 recombinants and finally 200 single clones), with autoradiographic screening at each stage, until a single cDNA clone was obtained which could transfer high capacity ¹²⁵I-hGM-CSF-binding to COS cells. Following three rounds of selection in COS cells, the individual cDNA clone in a second positive pool of DNA (pool 29) was ultimately identified as coding for the GM-CSF receptor by colony hybridization with the 1.8kbp insert of clone pGMR138. Plasmids encoding the cDNA inserts of clones 29 and 138 were designated pGMR29 and pGMR138 respectively.

EXAMPLE 4 Analysis of Cloned GM-CSF Receptor

The binding of [¹²⁵I]hGM-CSF to the transfected receptor on COS cells was specific, since it could be competed for by unlabelled hGM-CSF but not by murine GM-CSF (which has no activity on human cells), or by human IL-3, G-CSF or IL-6 (Fig. 3).

The binding characteristics of the cloned GM-CSF receptor (clone 138; Example 3) when transfected to COS cells are shown in Fig 4. The saturation binding isotherm for ¹²⁵I-hGM-CSF binding to transfected COS cells at 20°C showed a single class of binding site with an equilibrium dissociation constant of 6.8nM and 600,000 receptors per cell. In contrast, untransfected COS cells or COS cells transfected with vector alone showed no significant binding at these cell concentrations (3 to 7X10⁴ cells/point). Autoradiographic analyses indicated that only about 20% of the transfected cells were receptor positive (reflecting the transfection efficiency), so that positively transfected cells probably displayed about 3x10⁶ receptors per cell, compared with a reported value of 1700 receptors per cell for untransfected COS cells (Cocita Baldwin et al, 1989). Displacement of ¹²⁵I-hGM-CSF by unlabelled hGM-CSF also demonstrated only a single class of receptor with K_D=5.8nM, and this binding was not displaced by unlabelled hIL-3 (Fig 4B). The similar binding affinities observed when labelled or unlabelled hGM-CSF was varied indicates that iodination did not significantly alter the binding affinity of hGM-CSF for this receptor. In several different experiments binding of ¹²⁵I-hGM-CSF to transfected COS cells was determined in suspension (in binding medium with or without 20mM EDTA and 100 µg/ml chondroitin sulphate to prevent cell aggregation), or on adherent cells. The apparent K_D varied from 4 to 8nM, indicating that neither calcium nor adherence significantly altered the binding characteristics of the transfected receptor.

EXAMPLE 5 Molecular Size

The molecular size of the hGM-CSF receptor on HL60 cells and transfected COS cells was determined by chemical cross-linking with disuccinimidyl suberate(DSS) (Fig 5). As observed by others (DiPersio et al, 1988), the GM-CSF receptor on HL-60 cells had an approximate Mr of 85,000, so that cross-linking with ¹²⁵I-hGM-CSF (Mr=15,000) gave an Mr of 100,000. Both before and after rapid ligand dissociation (with or without low affinity binding, respectively - see Fig 1), a major cross-linked band of Mr 100,000 and a minor cross-linked band of Mr 95,000 (representing receptors of Mr 85,000 and 80,000, respectively) were seen; these may represent different glycosylation variants of a single binding subunit. Cross-linking of ¹²⁵I-hGM-CSF to the transfected receptor on COS cells gave a major band of similar molecular weight (90,000-110,000), although it was somewhat broader than that seen on HL-60 cells, possibly reflecting more variable glycosylation. Similar cross-linking gels run under reducing conditions gave the same cross-linked receptor molecular weight, indicating that the mature receptor was not comprised of disulphide-linked subunits.

EXAMPLE 6 Sequence Analysis

The inserts of clones 29 and 138 were subcloned and sequenced by standard techniques. In the composite sequence shown in Fig.6, clone 29 is represented by nucleotides 1-1709 and clone 138 by nucleotides 7-1807. The two sequences are otherwlse identical, except for a single silent base difference (G → A) in clone 29 at position 1148. Each sequence encodes a large open reading (ORF) frame of 400 amino acids preceded by a short ORF of 22 amino acids. The methionine codon beginning the larger reading frame is in a context which corresponds well to the consensus sequence (RCCATGG) for translation initiation sites (Kozak, 1987), while the shorter ORF begins with a methionine codon in poor context. The large ORF begins with a presumed signal peptide sequence of 22 amino acids, and residue Glu23 is assigned to the first amino acid of the mature protein by comparison with typical signal peptide cleavage sites (von Heijne, 1986).

The predicted 378 amino acid mature GM-CSF receptor is calculated to have a molecular weight of 43,728, which is approximately half the size of the receptor observed by cross linking of ¹²⁵I-hGM-CSF to HL60 cells and to COS-7 cells transfected with clone pGMR138 (Figure 5). This difference between the predicted size of the core receptor polypeptide and the mature receptor on cells is probably due to the attachment of carbohydrate to the 11 potential N-linked glycosylation sites in the putative 297 amino acid extracellular domain, since we have shown that the mature receptor does not contain disulphide-linked subunits. A hydrophobicity plot (Figure 7) suggests that a stretch of 27 uncharged amino acids extending from Gly320 to Phe346 (Figure 6) represents a transmembrane domain. This putative transmembrane domain is followed by a 554 amino acid intracellular domain which begins with a short stretch of basic amino acids, a feature common to the cytosolic face, next to the membrane-spanning segments, of many transmembrane proteins. The 54 amino acid intracellular domain of the GM-CSF receptor shares no apparent homology with that of the IL6-R.

In the 3'-untranslated region of the GM-CSF-receptor mRNA there is a sequence element homologous to the "Alu" family of repetitive elements (residues 1493 to 1760) which make up approximately 3% of the human genome (Jelinek and Schmid, 1982). Further downstream there are two poly A addition signals immediately prior to the poly A tail (Figure 6).

The predicted GM-CSF binding extracellular domain (amino acids 23-319) contains eleven cysteine residues, but these do not appear to form disulphide loops characteristic of receptors of the immunoglobulin superfamily (Sims et al, 1988; Yamasaki et al, 1988). The short intracellular domain (54 amino acids) may serve a role in signal transduction. This domain has no apparent sequence homology with the catalytic domain of any growth factor receptor known to be a tyrosine kinase (Hanks et al, 1988). However, direct comparison of the GM-CSF receptor with that of the human IL-6 receptor (Yamasaki et al, 1988) revealed significant homology (Fig. 9). The positions of four cysteine residues are approximately conserved (GM-CSF-R C₁₂₆, C₁₃₆, C₁₆₅, C₁₇₈ vs. IL6-R C₁₂₁, C₁₃₂, C₁₆₅, C₁₇₆), and these do not coincide with the two cysteine residues (C₄₇ and C₉₆) of the IL-6 receptor associated with the immunoglobulin-like domain (Yamasaki et al, 1988). Furthermore, there are patches of homology in the intracellular domain (Fig. 9) and a perhaps surprising similarity between the transmembrane domains of the two receptors (GM-CSF-R L₃₃- L₃₄₂ vs. IL6-R L₃₇₄-L₃₈₆), including in each case the transmembrane cysteine residue. The particular conservation of residues in the transmembrane regions may suggest a common ability to associate with other transmembrane proteins or lipids in their respective membranes. Indeed, a similarly placed cysteine residue in the transmembrane region of the Semliki Forest virus El spike protein, corresponding to the middle of the inner leaflet of the membrane, has been shown to be a site of palmitoylation (Schmidt et al, 1988). However, the position of the intramembrane cysteine residue relative to the putative inner face of the membrane (represented by R₃₈₈ (IL6-R) and K₃₄₇ (GM-CSF-R)) differs between the two receptors, so if this residue is functionally or structurally important it may,play a different role for each receptor.

The extracellular domain of the hGM-CSF receptor shows homology not only to that of the IL-6 receptor but also to those of the receptors for erythropoietin (D'Andrea et al, 1989), interleukin-2 (Hatekeyama et al, 1989), rat prolactin (Boutin et al, 1988) interleukin-4 (Mosley et al, 1989) and interleukin-3 (Itoh et al, 1990). The regions of homology include the four cysteine residues mentioned above and a short amino acid sequence centred around the sequence Trp-Ser-X-Trp-Ser near the transmembrane region (Fig. 9).

Prior to the long ORF encoding the GM-CSF receptor is a short ORF. This potentially encodes a polypeptide, although since its initiation methionine is in a context unrelated to the consensus sequence for good translation initiation (see above) this reading frame may not be translated (Kozak, 1986). Such short ORFs 5' of the main receptor coding region are found in the hIL6 receptor cDNA (Yamasaki et al, 1988), and one of these (25 nucleotides upstream) in fact begins with a methionine codon in a context that is stronger than that which was assigned to begin the hIL6 receptor precursor. This suggests that these short ORFs might act to depress the translation of the main receptor coding regions.

EXAMPLE 7 The GM-CSF Receptor Transcript

Since the cDNA clone pGMR138 was isolated from a human placental cDNA library, we tested whether or not mRNA corresponding to this transcript was also present in haemopoietic cells known to express the GM-CSF receptor.

Northern blot analysis (e.g. Fig. 8A) revealed that HL-60 cells, which are known to express high-affinity GM-CSF receptors, contain a 2.lkb transcript hybridizing at high stringency to the pGRM138 probe (tracks 3-5), whereas CEM T-lymphoblastoid cells and HepG2 hepatocellular carcinoma cells contain no detectable transcript hybridizing with this probe (tracks 1 and 2).

Owing to the low abundance of this RNA species, a more sensitive survey of RNA from various haemopoietic and non-haemopoietic cells was undertaken, using PCR-based amplification of cDNA corresponding to various RNAs. Such analyses (e.g. Fig 8B) revealed the presence of GM-CSF receptor transcripts in a variety of human myeloid cell lines, including HL-60, U937 and AML193, but not in CEM T-lymphoid, Raji Burkitt lymphoma nor Hep G2 hepatic cells. Interestingly we have found that HeLa cells also display GM-CSF receptors, and have transcripts corresponding to this cDNA clone (Figure 8B).

EXAMPLE 8 Introduction of hGM-CSF Receptor into Murine Cells

When the cloned low-affinity receptor for human GM-CSF from placental cells, described above, is introduced into a murine GM-CSF-dependent haemopoietic cell line (FDC-P1) using a retroviral vector, it retains its low-affinity phenotype but can nevertheless transmit the biological signals required for cell proliferation.

The murine FDC-Pl haemopoietic cell line used does not proliferate in cultures containing 10⁶ Units/ml of hGM-CSF, and no cells survive in such cultures. In four separate cocultivation experiments with 2 cells producing the hGM-CSF receptor retrovirus, 0.3 to 1% of FDC-P1 cells were able to proliferate clonally in agar cultures stimulated by hGM-CSF. Cloned cell lines (hGM-R-FD lines) were developed from individual colonies and maintained using either a high concentration of hGM-CSF, or a mixture of mGM-CSF with a lower concentration of hGM-CSF.

Southern blot analysis of DNA from 19 such hGM-R-FD lines (Figure 10a, tracks 2-20) revealed the presence of a single viral integrant in each clone, except clone 57 in which two integrants were evident (track 19). With the exception of clones 50 and 52, which are potentially siblings (tracks 13 and 14), the viral integration sites differed in each cloned line, as revealed by the different sizes of hybridizing DNA fragments, confirming the independent clonal origin of each line.

EXAMPLE 9 Differences between Cell Lines Maintained in Human GM-CSF versus Murine plus Human GM-CSF

At 25-39 days after the establishment of the cloned lines, a comparative analysis of 19 of these lines, chosen at random, revealed distinctive differences between the two types of lines in their behaviour in clonal cultures (Figure 11).

The parental FDC-P1 cell line usually exhibits a cloning efficiency in agar-medium of 60-100% when stimulated by mGM-CSF, forming large, tight, colonies. hGM-R-FD lines maintained in hGM-CSF usually exhibited lower clonogenic potential (42±17%), and total colony numbers were similar in cultures stimulated by h or mGM-CSF (Figure 11). The colonies characteristically had an irregular shape or were wholly dispersed, and maximal colony size was relatively small. Colony size was typically 2-4 times larger in parallel cultures stimulated by mGM-CSF. (For nine cell lines, mean colony size with mGM-CSF was 530±340 cells, versus 240±110 with hGM-CSF).

When hGM-R-FD cell lines maintained in murine plus human GM-CSF were stimulated by hGM-CSF, the frequency of clonogenic cells was even lower (15±15%) than that of lines maintained in hGM-CSF alone, although the morphology of the colonies was similar. With increasing duration of maintenance in the mixture of m+hGM-CSF, a progressive fall occurred in the frequency of clonogenic cells responsive to hGM-CSF alone. In sharp contrast, when cells of these lines were stimulated by mGM-CSF, the frequency of clonogenic cells was much higher (96±21%). The morphology and size of these colonies resembled those of parental FDC-Pl cells, and there was a more than tenfold difference between the size of colonies stimulated by m or hGM-CSF. For ten cell lines mean colony size with mGM-CSF was 1900±880 cells, versus 170±130 with hGM-CSF.

EXAMPLE 10 Responsiveness to GM-CSF of Lines Maintained inHuman versus Human plus Murine GM-CSF

The dose-response curves of hGM-R-FD lines responding to stimulation by h or mGM-CSF indicated that responsiveness to hGM-CSF was 500-1000-fold lower than to mGM-CSF (Figure 11, Table 1). Cell lines grown using hGM-CSF were twofold more responsive to hGM-CSF than cell lines maintained in a mixture of m+hGM-CSF, although both types exhibited a similar responsiveness to mGM-CSF (Figure 11, Table 1).

Quantitative Responsiveness of Cloned hGM-R FD Cell Lines to Stimulation by Murine or Human GM-CSF and Their Capacity to Bind Human GM-CSF.
Line No. CSF	Units Needed to Stimulate 50% Maximal Colonies	¹²⁵I-Human GM-Bound per 10⁶ Cells (cpm)
	mGM-CSF	hGM-CSF
Maintained in Murine plus Human GM-CSF
1	30	10,000	12,000
21	20	20,000	2,200
33	30	10,000	7,700
34	30	18,000	6,700
11	30	20,000	17,200
6	15	20,000	7,600
24	20	20,000	14,100
13	15	-	6,900
10	40	40,000	15,800
8	15	40,000	600
Maintained in Human GM-CSF Alone
54	20	10,000	9,300
53	20	10,000	9,000
58	7	7,000	11,800
49	20	10,000	73,400
50	15	<5,000	14,200
55	15	10,000	12,200
56	60	30,000	10,100
52	50	7,000	29,600
57	20	8,000	15,700
Control FDC-P1 Cells
	15	-	0
	15	-
	15	-

Three hundred cells were added to duplicate cultures containing increasing twofold concentrations of mouse or human GM-CSF. Colony counts were performed on day 7 and GM-CSF concentration stimulating 50% maximal colony numbers determined from each titration curve. Binding of ¹³¹I-labeled human GM-CSF determined in parallel using 5x10⁶ cells.

EXAMPLE 11 Display of hGM-CSF Receptors on Transfected FDC-P1 Cells

Different cloned hGM-R-FD lines were assessed for their capacity to specifically bind ¹²⁵I-hGM-CSF and, whilst all of them showed significant binding, there was considerable variation in the extent of this binding (Table 1). Clones maintained continuously in hGM-CSF alone showed higher average levels of bindlng than clones maintained in a mixture of hGM-CSF and mGM-CSF (Table 1).

Despite the variation in binding of ¹²⁵I-hGM-CSF to different hGM-R-FD clones, saturation binding analysis and Scatchard transformation of the binding data showed similar binding affinities (slopes of the Scatchard transformation, K_D=4-6nM) for all clones examined (Figure 12). This binding affinity was of a single low-affinity class and was the same as for receptors on human placental membranes, transfected COS-7 cells, and retrovirally-infected 2 clones.

The binding of hGM-CSF at concentrations up to 1 µg/ml to the transfected receptor in hGM-R-FD clones for 30 min at 37°C did not affect the subsequent binding of ¹²⁵I-mGM-CSF to the native mGM-CSF receptor co-expressed on these cells. Similarly, the binding of mGM-CSF or mMulti-CSF at concentrations up to 0.5 µg/ml to their native receptors on hGM-R-FD cells at 37°C did not affect the subsequent binding of ¹²⁵I-hGM-CSF to the transfected hGM-CSF receptor.

When ¹²⁵I-hGM-CSF was incubated with hGM-R-FD cells at 37°C it rapidly associated with cell surface GM-CSF receptors and was then slowly internalized into the cells. The binding and internalization kinetics were essentlally identical for hGM-R-FD clones maintained in m or hGM-CSF (Figure 13). However, the internalization rate of occupied hGM-CSF receptors (k_e) was slower in hGM-R-FD cells (k_e=0.0042 min^-1) than on human HL60 cells (k_e=0.061min^-1) and on occupied murine GM-CSF receptors on FDC-P1 cells (k_e=0.056 min^-1) (Nicola et al, 1988).

After retrovirally-mediated transfection of murine haemopoietic FDC-P1 cells with a cDNA encoding the low-affinity human placental GM-CSF receptor, cell-surface hGM-CSF receptors were displayed at a level of 10⁴-10⁵ per cell as a single binding class of low affinity (K_D=4-6nM). Despite careful analysis, no high-affinity binding was detected, yet the transfected cells could internalize the hGM-CSF receptor (albeit at a rate 10 times slower than for endogenous mGM-CSF receptors), and acquired the capacity to proliferate in response to hGM-CSF. The quantitative responsiveness of hGM-R-FD cells to hGM-CSF was 500-1000-fold lower than to mGM-CSF, but determination of the binding constants suggests that occupied hGM-CSF or mGM-CSF receptors might be equally efficient in transducing proliferative signals in murine FDC-Pl cells. Firstly, the transfected hGM-CSF receptor binds hGM-CSF with a 100-fold lower affinity (K_D=5nM) than that of the endogenous high-affinity receptor for mGM-CSF (K_D=50pM) (Walker and Burgess, 1985). Secondly, the 10-fold slower internalization rate of occupied hGM-CSF receptors compared to occupied mGM-CSF receptors at 37°C implies that the apparent steady-state "affinity constants" will differ by a further factor of 10 (Nicola et al, 1988).

EXAMPLE 12 Evolution of Recloned Sublines

Analysis of colonies grown from hGM-R-FD cell lines using m+hGM-CSF showed them to contain a major population of cells responsive only to mGM-CSF and a minor population responsive to hGM-CSF; this latter population rapidly diminished if colonies were grown for one week using only mGM-CSF. Colonies grown from cell lines maintained using hGM-CSF alone retained a stable content of clonogenic cells responding to either type of GM-CSF.

Cells from nine sublines derived from mGM-CSF-stimulated lines and maintained exclusively in mGM-CSF usually formed colonies only when stimulated by mGM-CSF (cloning efficiency 68±24%), and the colonies formed were of uniformly large size. Cells from 13 sublines, also derived from m+hGM-CSF-stimulated cultures but then maintained in human GM-CSF, formed relatively small numbers of colonies (cloning efficiency 37±25%) of medium size in cultures stimulated by h or mGM-CSF, with maintenance of the characteristic 2 to 4-fold size difference between the two types of colony.

In tests on six cloned sublines derived from hGM-CSF-stimulated lines and maintained in hGM-CSF, 50±26% of the clonogenic cells were able to form colonies in cultures containing 800 ng/ml antibiotic G418. In contrast, clonogenic cells from six cloned sublines maintained using mGM-CSF were uniformly unable to form colonies in the presence of G418. This suggested that transcription of the inserted neomycin resistance gene was not maintained when cells were stimulated solely by mGM-CSF.

Southern blot analysis of DNA from a series of mGM-CSF-responsive and human-responsive sublines of primary line 21 revealed that in all subclones both the content and context of the hGM-R viral integrant were maintained (eg. Figure 10A, tracks 21-27), confirming the common derivation of these lines, despite their divergent biological properties, and suggesting that the lack of both hGM-CSF responsiveness and display of cell surface hGM-CSF receptors was not due to loss of the hGM-R construct.

RNA from six sublines maintained with mGM-CSF contained no detectable hGM-R viral transcripts (Figure 10B, tracks 2-7), compared with abundant hGM-R viral transcripts evident in sublines maintained with hGM-CSF (tracks 8-11 and other data not shown), suggesting that the alteration in such cells was at the transcriptional or immediately post-transcriptional level.

Thus the behaviour of cloned hGM-R-FD cell lines differed, depending on whether they were maintained in hGM-CSF alone or a mixture of hGM-CSF and mGM-CSF. The former cell lines maintained a stable phenotype with equal clonogenicity in hGM-CSF or mGM-CSF, but clonogenicity was significantly lower than for cell lines maintained in mGM-CSF. The cell lines maintained in a mixture of mGM-CSF and a low concentration of hGM-CSF showed a progressive loss of cells able to respond to stimulation by hGM-CSF.

The behaviour of cell lines maintained in hGM-CSF was not influenced by the site of viral integration. The selection pressure exerted by culture with hGM-CSF alone maintained a constant level of expression of both viral genes (hGM-CSF-R and neoR), receptor expression and quantitative responsiveness to hGM-CSF. However, the low clonogenicity and colony size exhibited by these cells suggested the continuous generation within lines maintained in hGM-CSF of progeny cells that had lost responsiveness to hGM-CSF. In the absence of any other stimulus, these cells irreversibly lost proliferative potential and could not be rescued by subsequent culture in mGM-CSF. This phenomenon was confirmed by analysis of subcloned hGM-R-FD lines maintained in the presence of mGM-CSF only. Such subcloned lines maintained their viral inserts, but expression of both hGM-CSF receptor and neoR genes declined progressively to zero with increasing time in culture. The reasons for the relatively high frequency at which the viral genes were turned off are unclear, but this clearly did not depend on viral integration site. Indeed, suppression of retroviral expression has been observed in a variety of different cells, including murine haemopoietic cells (Chang et al, 1987; Magli et al, 1987; Emerman and Temin, 1984); the SV40 early region promoter, used in the present construct to drive the neomycin resistance gene, has been shown to be particularly sensitive to trans- and cis-acting negative regulatory factors (Chang et al, 1987; Magli et al, 1987; Emerman and Temin, 1984; Gorman et al, 1985; Williams et al, 1986).

The observation that the human placental GM-CSF receptor stimulated the proliferation of murine haemopoietic FDC-Pl cells lends support to the suggestion that this receptor subunit may form a component of the human GM-CSF receptor on haemopoietic cells. It also demonstrates that a low affinity GM-CSF receptor of non-haemopoietic origin can transduce proliferative signals and be internalized in a haemopoietic cell, in a ligand-dependent manner, in the absence of any high-affinity binding component. Indeed, since biological responsiveness of the transfected FDC-P1 cells was almost the same to m or hGM-CSF on a receptor-occupied basis, this further suggests that the signalling components of both h and mGM-CSF receptors may be highly conserved, and that the main function of high affinity subunits might be solely to increase the rate of receptor internalization and to increase the responsiveness of haemopoietic cells to low external concentrations of GM-CSF. Alternative interpretations of the data involving interaction or cointernalization of the introduced hGM-CSF receptor with endogenous mGM-CSF or Multi-CSF receptors were excluded.

Despite the conservation of signalling between human and murine GM-CSF receptors, there is no functional conservation in the GM-CSF-binding domain, and there was no conversion of the hGM-CSF receptor to a high-affinity receptor by interaction with the mouse cells that displayed high-affinity mGM-CSF receptors. The relationship of the cloned low-affinity hGM-CSF receptor to high-affinity GM-CSF receptors remains unclear. It is still possible that high-affinity GM-CSF receptors are unrelated to the low-affinity hGM-CSF receptor. However, if, as we have postulated, the low affinity subunit can be converted to a high-affinity receptor by interaction with an "adaptor" subunit, then this interaction seems not to occur across these species.

A wide variety of potential therapeutic, diagnostic and preparative applications of GM-CSF receptors of this invention is envisaged, including, but not limited to, the following:

1. Inhibition of the growth of GM-CSF dependent myeloid leukaemias.

2. Treatment of patients who have been overdosed with GM-CSF.

3. Local treatment of adverse side-reactions in patients receiving GM-CSF.

4. Control of systemic or local GM-CSF levels in patient with over-exuberant or inappropriate inflammatory responses.

5. Reduction of inflammatory response in patients with chronic infections (eg. pulmonary fungal infections, listeriosis, tuberculosis, acute respiratory distress syndrome), autoimmune reactions or inappropriate GM-CSF production.

6. Classification of myeloid leukaemias for stratification for outcome and selection for therapy.

7. Detection of aberrant expression of GM-CSF-R in tumours to allow selection of patients to receive combination therapy with GM-CSF and chemotherapy (exclusion of cancers with GM-CSF-R) (eg. lung, breast and bladder cancer, myeloid leukaemias).

8. Screening of patients with capacity to respond to GM-CSF in aplastic anaemias and congenital neutropaenias.

9. Identification of disease states mediated by autoimmune anti-GM-CSF-R antibodies (eg. autoimmune neutropaenias).

10. Preparation of affinity matrices to allow rapid purification of GM-CSF for clinical studies and therapeutic use.

11. Screening of potential agonists and antagonists of GM-CSF action.

12. Preparation of antibodies against the hGM-CSF-R to be used in:

a. identification and quantitation of circulating soluble GM-CSF-R,

b. screening marrow cells for evaluation prior to therapy

c. evaluating marrow reserve in pre-treatment patients,

d. determining pharmacokinetics of GM-CSF-R in clinical trials of GM-CSF-R.

13. Preparation of anti-idiotypic antibodies to mimic the action of GM-CSF for use in therapy and in lowering systemic lipid levels.

14. Preparation of nucleic acid probes for identification of aberrant GM-CSF receptor genes in disorders such as myeloid leukaemia.

Although these applications are listed with reference to hGM-CSF, it will be apparent to the person skilled in the art that corresponding veterinary diagnostic, therapeutic and preparative applications of appropriate animal GM-CSF-Rs are within the scope of the invention.

References cited herein are listed on the following pages.

It will be clearly understood that the invention in its general aspects is not limited to the specific details refered to hereinabove.

REFERENCES

1. Aruffo, A.; Seed, B. (1987): Proc. Natl. Acad. Sci. (USA), 84, 8573-8577.

2. Boyd, A.; Wawryk, S.O.; Burns, G.F.; Fecondo, J.V. (1988): Proc. Natl. Acad. Sci. (USA), 85, 3095-3099.

3. Boutin, J. M.; Jolicoeur, C.; Okamura, H.; Gagnon, J.; Edery, M.; Shirota, M.; Banville, D.; Dusanter-Fourt, I.; Djiane, J.; Kelly, P.A. (1988): Cell, 53, 69-77.

4. Bussolino, F.; Wang, J.M.; Defilippi, P.; Turrini, F.; Sanavio, F.; Edgell, C-J.S.; Aglietta, M.; Arese, P.; Mantovani, A. (1989): Nature, 337, 471-473.

5. Calvo, J.C.; Radicella, J.P.; Charreau, E.H. (1983): Biochem.J., 212, 259-264.

6. Cepko, C.L.; Roberts, B.E.; Mulligan, R.C. (1984): Cell,37, 1053-1062.

7. Chang, J.M.W.; Wager-Smith, K.; Tsao, T.Y.; Tigges, J.H.; Vaishnav, S.; Caskey, C.T. (1987): Mol. Cell. Biol., 7, 854-863.

8. Cocita Baldwin, G.; Gasson, J.C.; Kaufman, S.E.; Quan, S.G.; Williams, R.E.; Avalos, B.R.; Gazdar, A.F.; Golde, D.W.; Di Persio, J.F. (1989): Blood, 73, 1033-1037.

9. D'Andrea, A.D.; Lodish, M.F.; Wong, G.G. (1989): cell, 57, 277-285.

10. Dedhar, S.; Gabourg, L.; Galloway, P.; Eaves, C. (1988): Proc. Natl. Acad. Sci. (USA), 85, 9253-9257.

11. Delamarter, J.F.; Mermod, J.J.; Liang, C.M.; Eliason, J.F.; Thatcher, D.R. (1985): EMBO. J., 4, 2575-2581.

12. Dexter, T.M.; Garland, J.; Scott, D.; Scolnick. E.; Metcalf, D. (1980): J. Exp. Med., 152, 1036-1047.

13. Di Persio, J.; Billing, P.; Kaufman, S.; Eghtesady, P.; Williams, R.E.; Gasson, J.C. (1988): J. Biol. Chem., 263, 1834-1840.

14. Donohue, R.E.; Wang, E.A.; Stone, D.K.; Kamen, R.; Wong, G.G.; Sehgal, P.K.; Nathan, D.G.; Clark, S.C. (1986): Nature, 321, 827-875.

15. Elliot, M.J.; Vadas, M.A.; Eglinton, J.M.; Park, L.S.; To, L.B.; Cleland, L.G.; Clark, S.C.; Lopez, A.F. (1989) Blood, 74, 2349-2359.

16. Emerman, M.; Temin, H.M. (1984): Cell, 39, 459-467.

17. Evans, D.B.; Bunning, R.A.D.; Russell, R.G.G. (1989): Biochem. Biophys. Res. Commun., 160, 588-595.

18. Feinberg, A.P.; Vogelstein, B. (1983): Anal. Biochem., 132, 6-13.

19. Gasson, J.C.; Kaufman, S.E.; Weisbart, R.H.; Tomonaga, M.; Golde, D.W. (1986): Proc. Natl. Acad. Sci. (USA), 83, 669-673.

20. Gesner, T.G.; Mufson, R.A.; Norton, C.R.; Turner, K.J.; Yang, Y-C.; Clark, S.C. (1988): J. Cell. Physiol., 136, 493-499.

21. Gorman, C.M.; Rigby, P.W.S.; Lane, D.P. (1985): Cell; 42, 519-526.

22. Gough, N.M. (1988): Anal. Biochem., 173, 93-95.

23. Gough, N.M. and Nicola, N.A. (1989): In "Colony-Stimulating Factors: Molecular and Cellular Biology", (Hexter, T.M., Garland, J. and Testa, N. eds): Marcel Dekker, N.Y., 111-153.

24. Gough, N.M.; Gough, J.; Metcalf, D.; Kelso, A.; Grail, D.; Nicola, N.A.; Burgess, A.W.; Dunn, A.R. (1984): Nature, 309, 763-767.

25. Hanks, S.K.; Quinn, A.M.; Hunter, T. (1988): Science, 241, 42-52.

26. Hatekeyama, M.; Tsudo, M.; Minamoto, S.; Kono, T.; Miyata, T.; Miyasaka, M.; Taniguchi, T. (1989): Science, 244, 551-556.

27. Hilton, D.J.; Nicola, N.A.; Metcalf, D. (1989): Proc. Natl. Acad. Sci. (USA), 85, 5971-5975.

28. Hopp, T.P. and Woods, K.R. (1983): Mol. Immunol., 20, 483-489.

29. Itoh, N.; Yonehara, S.; Schreurs, J.; Gorman, D.M.; Maruyama, K.; Ishii, A., Yahara, I.; Arai, K.-I.; Miyajima, A. (1990): Science, 247, 324-327.

30. Jelinek, W.R. and Schmid, S.W. (1982): Ann. Rev. Biochem., 51, 813-844.

31. Johnson, G.R.; Gonda, T.J.; Metcalf, D.; Hariharan, T.K.; Cory, S. (1989): EMBO J., 8, 441-448.

32. Kelleher, C.A.; Wong, G.G.; Clark, S.C.; Schendel, P.F.; Minden, M.D.; McCulloch, E.A. (1988): Leukaemia, 2, 211-215.

33. Korman, A.J.; Frantz, J.D.; Strominger, J.L.; Mulligan, RC. (1987): Proc. Natl. Acad. Sci. (USA), 84, 2150-2154.

34. Kozak, M. (1986): Cell, 44, 283-292.

35. Kozak, M. (1987): Nucl. Acids Res., 12, 857-872.

36. Lang, R.A.; Metcalf, D.; Cuthbertson, R.A.; Lyons, I.; Stanley, E.; Kelso, A.; Kannourakis, G.; Williamson, D.J.; Klintworth, G.K.; Gonda, T.J.; Dunn, A.R. (1987): Cell, 51, 675-686.

37. Lopez, A.F.; Eglington, J.M.; Gillis, D.; Park, L.S.; Clark, S.; Vadas, M.A. (1989): Proc. Natl. Acad. Sci. (USA), 86, 7022-7026.

38. Magli, M-C.; Dick, J.E.; Huszar, D.; Bernstein, A.; Phillips, R.A. (1987): Proc. Natl. Acad. Sci. (USA), 84, 789-793.

39. Maniatis, T.; Fritsch, E.F.; Sambrook, J. (1982): Molecular Cloning - A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 1-545.

40. Mann, R.; Mulligan, R.C.; Baltimore, D. (1983): Cell, 33, 153-159.

41. Metcalf, D. (1984): The Hemopoietic Colony Stimulating Factors, Elsevier, Amsterdam.

42. Metcalf, D.; Johnson, G.R.; Burgess, A.W. (1980): Blood, 55, 138-147.

43. Metcalf, D.; Begley, C.G.; Williamson, D.J.; Nice, E.C.; DeLamarter, J.; Mermod, J-J.; Thatcher, D.; Schmidt, A. (1987): Exp. Hematol., 15, 1-9.

44. Morstyn, G.; Lieschke, G.J.; Sheridan, W.; Layton, J.; Cebon, J. (1989): Trends Pharmacol. Sci., 10, 154-159.

45. Mosley, B.; Beckmann, M.P.; March, C.J.; Idzerda, R.L.; Gimpel, S.D.; Vanden Bos, T.; Friend, D.; Alpert, A.; Anderson, D.; Jackson, J.; Wignall, J.M.; Smith, C.; Gallis, B.; Sims, J.E.; Urdal, D.; Widmer, M.B.; Cosman, D.; Park, L.S. (1989): Cell, 59, 335-348.

46. Munson, P.J. and Robard, D. (1980): Anal. Biochem., 107, 220-239.

47. Nicola, N.A. (1987): Immunol. Today, 8, 134-139.

48. Nicola, N.A. and Metcalf, D. (1985): J. Cell. Physiol., 124, 313-321.

49. Nicola, N.A. and Peterson, L. (1986): J. Biol. Chem., 261, 12384-12389.

50. Nicola, N.A.; Peterson, L.; Hilton, D.J.; Metcalf, D. (1988): Growth Factors, 1, 41-49.

51. Nikaido, T.; Shimuzu, A.; Ishida, N.; Sabe, H.; Teshigawara, K.; Maeda, M.; Uchiyama, T.; Yodoi, J.; Honjo, T. (1984): Nature, 311, 631-635.

52. Padmanabhan, R.; Corsico, C.D.; Howard, T.H.; Holter, W.; Fordis, C.M.; Willingham, M.; Howard, B.M. (1988): Anal. Biochem., 170, 341-348.

53. Park, L.L.; Friend, D.; Gillis, S.; Urdal, D.L. (1986a): J. Biol. Chem., 261, 4177-4183.

54. Park, L.L.; Friend, D.; Gillis, S.; Urdal, D.L. (1986b): J. Exp. Med., 164, 251-262.

55. Park, L.S.; Friend, D.; Price, V.; Andereson, D.; Singer, J.; Prickett, K.S.; Urdal, D.L. (1989): J. Biol. Chem., 264, 5420-5427.

56. Rettenmier, C.W.; Sacca, R.; Furman, W.L.; Roussel, M.F.; Holt, J.T.; Nienhuis, A.W.; Stanley, E.R.; Sherr, C.J. (1986): J. Clin. Invest., 77, 1740-1746.

57. Sanger, F.; Nicklen, S.; Coulson, A.R. (1977): Proc. Natl. Acad. Sci. (USA), 74, 5463-5467.

58. Scatchard, G. (1949): Ann. N.Y. Acad. Sci., 51, 660-672.

59. Schmidt, M.; Schmidt, M.F.G.; Roff, R. (1988): J. Biol. Chem., 263, 18635-18639.

60. Simmons, D.; Makgoba, M.W.; Seed, B. (1988): Nature, 331, 624-627.

61. Sims, J.E.; March, C.J.; Cosman, D.; Widmer, M.B.; MacDonald, H.R.; McMahan, C.J.; Grabon, C.E.; Wignall, J.M.; Jackson, J.L.; Call, S.M.; Friend, D.; Alpert, A.R.; Gillis, S.; Urdal, D.L.; Dower, S.K. (1988): Science, 241, 585-589.

62. Tabor, S. and Richardson, C.C. (1987): Proc. Natl. Acad. Sci. (USA), 84, 4767-4771.

63. Von Heijne, G. (1986): Nucl. Acids Res., 14, 4683-4690.

64. Wegmann, T.G.; Athanassakis, I.; Guilbert, L.; Branch, D.; Dy, M.; Menu, E.; Chaouat, G. (1989): Transplant. Proc., 21, 566-568.

65. Walker, F. and Burgess, A.W. (1985): EMBO J., 4, 933-939.

66. Walker, F.; Nicola, N.A.; Metcalf, D.; Burgess, A.W. (1985): Cell, 45, 269-276.

67. Williams, D.A.; Orkin, S.H.; Mulligan, R.C. (1986): Proc. Natl. Acad. Sci., (USA), 83, 2566-2570.

68. Wong, G.G.; Witek, J.S.; Temple, P.A.; Wilkens, K.M.; Leary, A.G.; Luxenberg, D.P.; Jones, S.S.; Brown, E.L.; Kay, R.M.; Orr, E.C.; Shoemaker, C.; Golde, D.W.; Kaufmann, R.J.; Hewick, R.M.; Wang, E.A.; Clark, S.C. (1985): Science, 228, 810-.

69. Yamasaki, K.; Hirata, Y.; Yawara, M.; Kawanishi, Y.; Seed, B.; Taniguchi, T.; Hirano, T.; Kishimoto, T. (1988): Science, 241, 825-828.

70. Yeung, Y.G.; Jubinsky, P.T.; Sengupta, A.; Yeung, D.C.; Stanley, E.R. (1987): Proc. Natl. Acad. Sci. (USA), 84, 1268-1271.

71. Young, Y.G. and Griffin, J.D. (1986): Blood, 68, 1178-1181.

Claims

A purified nucleic acid selected from the group consisting of single stranded DNA, double stranded DNA, cDNA and RNA which each comprises a nucleotide sequence which encodes, or is complementary to a nucleotide sequence which encodes, human GM-CSF receptor or a derivative thereof capable of binding human GM-CSF, said GM-CSF receptor being defined as a glycosylated or unglycosylated proteinaceous molecule comprising, in its entirety, an extracellular domain, a transmembrane domain, and an intracytoplasmic tail and being capable of specifically binding radioactively labelled GM-CSF or derivatives thereof, said binding being competed for by unlabelled GM-CSF.
A nucleic acid according to Claim 1, wherein said derivative has at least 75% amino acid homology with naturally occurring GM-CSF receptor.
A nucleic acid according to Claim 1 or claim 2, wherein the nucleic acid is a DNA and comprises a nucleotide sequence which is as set out in Figure 6B or which has at least 75% homology with said sequence of Figure 6B.
A nucleic acid according to any one of Claims 1 to 3 which additionally comprises a further nucleic acid sequence contiguous to the sequence encoding human GM-CSF receptor.
A nucleic acid according to any one of Claims 1 to 4 wherein said GM-CSF receptor comprises an extracellular domain which has the capacity to bind GM-CSF.
A recombinant DNA molecule comprising a gene encoding a human GM-CSF receptor which has the amino acid sequence as set out in Figure 6B, or a homologue thereof which encodes a polypeptide having at least 75% sequence identity with the said receptor.
A recombinant DNA molecule according to Claim 6, said polypeptide comprising an extracellular domain, a transmembrane domain, and an intracellular domain.
A plasmid or replicable vector comprising a nucleic acid as defined in any one of Claims 1 to 5 or recombinant DNA molecule according to Claim 6 or Claim 7.
A recombinant cell comprising a nucleic acid as defined in any one of Claims 1 to 5, or a recombinant DNA molecule according to claim 6 or claim 7.
A recombinant cell according to Claim 9 which is a simian COS cell.
A recombinant cell according to Claim 9 which is a haemopoietic cell, wherein the nucleic acid encodes a receptor for a heterologous GM-CSF.
A recombinant cell according to Claim 11 which is a GM-CSF dependent haemopoietic cell.
Recombinant or synthetic human GM-CSF receptor or a derivative thereof capable of binding human GM-CSF, essentially free of other proteins, said GM-CSF receptor being defined as a glycosylated or unglycosylated proteinaceous molecule comprising, in its entirety, an extracellular domain, a transmembrane domain, and an intracytoplasmic tail and being capable of specifically binding radioactively labelled GM-CSF or derivatives thereof, said binding being competed for by unlabelled GM-CSF.
Recombinant or synthetic human GM-CSF receptor according to Claim 13, wherein said derivative has at least 75% amino acid homology with naturally occurring GM-CSF receptor.
A recombinant or synthetic GM-CSF receptor according to Claim 13 which comprises any of the amino acid sequences set out in Figure 6B or which has at least 75% homology with any said amino acid sequence of Figure 6B.
A recombinant or synthetic GM-CSF receptor according to any one of Claims 13 to 15 which is a fusion protein or a hybrid protein.
Recombinant or synthetic GM-CSF receptor according to any one of Claims 13 to 16, wherein said GM-CSF receptor comprises the extracellular domain which has the capacity to bind GM-CSF.
GM-CSF receptor produced by a recombinant cell as defined in any one of Claims 9 to 12.
An antibody or an antigen-binding fragment thereof which antibody or fragment specifically binds a recombinant or synthetic GM-CSF receptor according to any one of Claims 13 to 18.
A pharmaceutical composition comprising a recombinant or synthetic GM-CSF receptor according to any one of Claims 13 to 18, or an antibody or an antigen-binding fragment thereof according to Claim 19.
Use of recombinant or synthetic GM-CSF receptor or its derivative, as defined in any one of Claims 13 to 18, or of an antibody or antigen-binding fragment thereof as defined in Claim 19 for the manufacture of a medicament for use in the treatment of a GM-CSF related diseases.
Use of GM-CSF receptor or its derivative or an antibody or antigen-binding fragment thereof as defined in Claim 19 for the manufacture of a composition for use in modulating the proliferation, differentiation or functional activation of GM-CSF-stimulation-sensitive cells in a mammal.
A method of generating information for use in the diagnosis of a GM-CSF related disease in a mammal, comprising the step of contacting a tissue or biological fluid or said mammal, in vitro, with a nucleic acid or recombinant DNA molecule as defined in any one of Claims 1 to 7, a recombinant cell as defined in any one of Claims 9 to 12, a recombinant or synthetic GM-CSF receptor as defined in any one of Claims 13 to 18, or an antibody or antigen-binding fragment thereof as defined in Claim 19 and detecting the thus-formed product.
Use of a nucleic acid or recombinant DNA molecule as defined in any one of Claims 1 to 7, a recombinant cell as defined in any one of Claims 9 to 12, a recombinant or synthetic GM-CSF receptor as defined in any one of Claims 13 to 18, or an antibody or antigen-binding fragment thereof as defined in Claim 19 in the preparation of a composition for use in the diagnosis of a GM-CSF related disease in a mammal.
A method according to Claim 23 or use according to claim 24, wherein said GM-CSF related disease is a cancer, tumour or leukaemia caused by or associated with cells sensitive to stimulation by GM-CSF.
A method of screening a cDNA library for a cDNA fragment encoding GM-CSF receptor wherein said GM-CSF receptor is defined as a glycosylated or unglycosylated proteinaceous molecule comprising, in its entirety, an extracellular domain, a transmembrane domain, and an intracytoplasmic tail and being capable of specifically binding radioactively labelled GM-CSF or derivatives thereof, said binding being competed for by unlabelled GM-CSF, comprising the steps of:

(a) constructing a cDNA library,

(b) preparing cDNA fragments therefrom,

(c) transfecting said fragments into mammalian host cells,

(d) incubating said transfected cells with labelled GM-CSF,

(e) identifying populations of transfected cells binding said labelled GM-CSF,

(f) preparing clones of host cells transformed with cDNA fragments able to cause mammalian host cells to bind labelled GM-CSF, and isolating said clones.
A process for production of GM-CSF receptor comprising the steps of culturing a recombinent cell according to any one of Claims 9 to 12 in a suitable nutrient medium, and recovering GM-CSF receptor.
A method of preparing a DNA sequence encoding GM-CSF receptor wherein said GM-CSF receptor is defined as a glycosylated or unglycosylated proteinaceous molecule comprising, in its entirety, an extracellular domain, a transmembrane domain, and an intracytoplasmic tail and being capable of specifically binding radioactively labelled GM-CSF or derivatives thereof, said binding being competed for by unlabelled GM-CSF, comprising the steps of:

(a) providing a cDNA library from a source which expresses GM-CSF receptor;

(b) preparing cDNA fragments therefrom;

(c) transfecting said fragments into host cells;

(d) identifying host cells which express GM-CSF receptor;

(e) recovering the host cells which express GM-CSF receptor, and

(f) recovering DNA from the host cells recovered in step (e).
A method according to Claim 28 in which step (d) is carried out by the method of Claim 26.
A method according to any one of Claims 26, 28, or 29, in which the cDNA library is derived from human placenta.
Plasmids pGMR138 and pGMR29 as defined in Example 3 herein.
Cell line hGM-R-FD as defined in Example 8 herein.
A method of screening of compounds for ability to augment or reduce binding of GM-CSF, comprising the steps of contacting a cell line bearing recombinant GM-CSF receptors with GM-CSF labelled with a detectable marker, in the presence of compound to be tested, and measuring whether binding of GM-CSF to the transfected cells is augmented or reduced compared to binding of a reference compound, wherein said GM-CSF receptor is defined as a glycosylated or unglycosylated proteinaceous molecule comprising, in its entirety, an extracellular domain, a transmembrane domain, and an intracytoplasmic tail and being capable of specifically binding radioactively labelled GM-CSF or derivatives thereof, said binding being competed for by unlabelled GM-CSF.
A method according to Claim 33 wherein the cell line is hGM-R-FD as defined in Example 8 herein.
A nucleic acid probe whose sequence is comprised within the sequence set out in Figure 6B, or is complementary thereto.